EMI generated from my Eigantakt Class D Amplifier
- Thread starter bornie
- Start date
Not so sure about this. Just like all digital induces a digital bi-directional noise, my limited experience with only a few Class D amp mfg’ers tells me that all Class D will likewise induce digital-like bi-directional noise. So it’s not so much whether one decrees it pretty low noise or not, it’s a given that all Class D amps induce this bi-directional noise and if the channels share the same AC inlet, it’s guaranteed they will share that noise between channel.
What’s not so given is whether or not this Class D distortion by-product has been sufficiently addressed, right? Isn’t that the real question?
I’m aware of Bruno, his reputation, and the accolades that follow him. But that’s not a guarantee that matters have been sufficiently addressed.
I’m also aware that simply believing one’s word is never enough. High-end audio dogma, listening skills, and talents are all over the map with few if any standards and all too often we put too much faith in others without confirmation.
Anyway, it sounds like we're in agreement that all Class D induces a bi-directional digital-like noise. And it sounds like you put some serious due diligence into your correction(s).
So let me ask you…. Do your own Class D stereo amps or even Bruno’s include dedicated AC inlets - one for each channel?
If not, considering we live in an imperfect world, wouldn't it still be best practice to design in dedicated AC inlets just to ensure any bi-directional noise overlooked or remaining does not interchange between channels? Wouldn’t such a minor precaution at least guarantee this bi-directional distortion is not shared / multiplied between channels and put this very common problem to rest once and for all?
Especially for the integrated flavors that must borrow and then add gain from one of these potentially already polluted channels?
Gotta' remember that the OP was complaining about EMI - or at least some type of distortion.
Last edited:
It seems you are correct in that you are not sure about this.
If you have a small sample size and then assume that all things are like that small sample, then its very easy to mislead yourself.
Short answer is 'no'. The longer answer is very similar, longer because of the explanation (already given above) as to why.
That is why I said "I see it as highly unlikely one of his designs would have this problem. All bets are off if there's a malfunction."
Our class D amps are monoblocks. The prototypes were not and had to play nice on the same chassis. That happened because we made sure the switching noise was well below the 'recommended minimums'.
No, it wouldn't. If there's a noise problem, hetrodyning noise will show up whether separate power cords are used or not!!
lol. Earlier you acknowledged and agreed that bi-directional noise was a universal problem with Class D but now you’re singing a different tune?
First you downplay the significance of Class D’s universal bi-directional digital-like noise and now you claim it’s only an issue with the two Class D mfg’ers I’ve experience with. Thereby, ignoring anybody else who’s expressed similar concerns with their Class D amp’s bi-directional noise?
Interesting. Whose recommended minimums?
Couldn’t one easily interpret your response as… You initially prototyped Class D stereo amp chassis’ but the 2 channels could not “play nice” sharing the same chassis i.e. same AC inlet? Why else might a designer entirely abandon their resource-consuming stereo chassis prototype designs and testing and instead go the monoblock route?
Wasn’t Class D monoblocks (or dual AC inlets at the stereo chassis) my point from the git-go?
Assuming of course the designer is measuring and listening for the right things, right?
And all the while we’ve still got the OP with a Class D integrated amp reaching out for answers regarding some Class D int. amp distortions he cannot overlook. Distortions that you claim must be a malfuction because in your opinion, this universal Class D bi-directional noise has either been absolutely minimized or resolved by both you and Bruno?
Sorry, but the more you defend, the more my radar goes up.
No. I don't think you read my post properly at that time.
I immediately translated your 'bidirectional noise' phrase to simply mean switching noise. All noise, FWIW is 'bidirectional'; anything not DC is 'bidirectional' as it is AC in nature.
One recommended minimum to which every manufacturer has to pay attention is the EU directives that one must meet in order to acquire the CE mark for import into the EU.
One could, and would be wrong in doing so. Switching noise due to its high frequency nature is just like RFI (it is literally the same thing) in that its insidious. It can radiate through the air, through ground connections, through various bits of wire and certainly does not need to be in the same box to be a problem!
Monoblocks are preferred if you want to get the most out of your system. This is because they can be placed as close to the speaker as possible, thereby keeping speaker cables short. Long speaker cables, especially if the speakers are 4 Ohms (or less), rob the system of bass impact and resolution. I don't like putting my equipment stand between the speakers as that means the front end of the system will be subjected to more energy from the loudspeakers. So I like to place it elsewhere and run long (balanced) interconnects to the amps; since our equipment supports AES48 we can run interconnect cables 100 feet without any coloration or degradation.
Monoblocks allow this; a stereo chassis must be placed between the speakers to keep speaker cables short so will always be less than optimum. This is the reason most of our tube amplifiers are monoblocks too.
Yes. You simply drew the wrong conclusion for why they might exist. Its not to eliminate problems of two channels in the same box as that would never work. If the two modules can't play nice in such a situation you have a serious problem; either a malfunction or a design flaw. Its occurring to me that you might not understand how quiet a class D amplifier actually has to be in order to work right. If its got a problem, the fact that its in one box or two will not prevent noise from getting into the system and being a problem.
Designing a class D amplifier requires a bit of math and a lot of measurement; IME quite a lot more than you might have to go through designing a class A or AB amplifier whether tube or solid state. In addition the equipment needed to make the measurements needed is considerably more sophisticated than you might need for the prior art. IOW anyone who has designed a functional class D module did not do it without a lot of measurement. Your question is a bit of a red herring.
I'm simply trying to get you to understand some nuance of how class D works since your initial post to this thread indicated a misunderstanding. For example your phrase 'bidirectional noise' really isn't a thing. Its just switching noise as I explained earlier. You've been laboring under a false impression. I've simply been attempting to disabuse you of that, perhaps too gently.
You do realize that I don't have a dog in the fight? Whatever is going on with the OP's system, its obviously a malfunction (which might be fixed, since he is absent from this thread at this point). If it was a problem with two modules in the same box, the noise would be constant.
Now you could ask someone who never designed a class D module and see how much accurate information you get. But if you ask someone (and I admit, most class D designers are not active on audio forums like this so that makes this difficult) who has and you'll get answers like mine.
Clearly we’ll just have to agree to disagree on terminology, significance, and potential remedies.
Regardless of what anybody calls it, e.g. switching noise, RFI, EMI, or bi-directional noise, etc., bi-directional noise is a real byproduct of Class D operation that can travel back through the shared AC inlet. That’s not in dispute; even excellent engineering can only reduce it, not make it vanish entirely.
And while most noise / distortions are fundamentally technically AC as Atmosphere pointed out, the kind of high-frequency switching hash a Class D amp throws off between channels and linestage is far more invasive than the usually more gentle distortions we’re used to from analog gear as it can leak back into the system through shared AC inlet in a way that really does matter.
It’s also worth noting that without superior passive, dedicated, bi-directional filtering line conditioners this noise can leak to all other components even if they are on separate circuits as some will claim this digital-like distortion can make its way back to the service panel and induce its harm on other circuits too.
FWIW, years ago I ran a few simple experiments that made this clear to me.
Around 2006, I wired the IEC connector of one of my treasured Foundation Research LC 10 passive, dedicated, bi-directional line conditioner to service a pair of NuForce monoblocks, and the added distortions were immediately audible.
I repeated a similar test around 2015 with my current Wyred-4-Sound (W4S) monoblocks and got the same results.
Just prior to taking possession of my W4S monoblocks, I had purchased a Wyred-4-Sound Class D integrated amp (single AC inlet) and the outcome was far worse. The presentation was horrifically unmusical so I swapped it out for the very musical W4S monoblocks I still use today.
Whether or not this kind of noise is audible will always depend on the design, the system, and the listener. My only point here is to inform the OP, given their integrated design, that if you hear something that smells like Class D switching noise, you’re probably not hearing a malfunction, but most likely encountering real bi-directional noise compounding between channels and the linestage. Which, I suppose, could be considered a malfunction if the amp designer claimed to have solved it (like Atmosphere) but it’s still prevalent.
And if per chance the culprit is the bi-directional digital-like noise I suspect, the only reasonable mitigation I’m aware of is to move to Class D monoblocks AND superior line conditioners (highly recommend Jena Labs series of line conditioners).
Many claim Class D amps are quite musical, while many others report hearing, excessive brightness or detail, or digital-like hash i.e. bi-directional noise. That’s where things like monoblocks or dual AC inlets, and superior line conditioning can help mitigate the damage.
That said, I can’t help but think… With Class D’s musical potential even in its current state, if someone had genuinely absolutely minimized this rather obvious bi-directional noise (as Atmosphere claims he has), it would have made breaking headline news by now and far more people would be swearing allegiance to Class D amplification. As of yet, I’m unaware that either has happened.
You know I make sound suggestions based on facts no reply
I wonder is it just the trash a product you no longer like or did not like but can’t get a refund
let ralph explain but it can’t be EMI had to be RFI
BE it power or network or anything in between
EMI is magnetic has you thought into how much is needed and how distance makes it become minute in a few feet
tubes can be effected by both but still needs to be close enough if EMI
if there is a router with enabled WiFi even 10 feet away or on another floor or room it can effect things. this is a proven fact and has happened to me
Also most all devices have switching psu now and this includes linaer power supplies
a regulator is also switching as is diodes
So get off the soap box of how you absolutely know it’s your disliked class d amp
I wonder is it just the trash a product you no longer like or did not like but can’t get a refund
let ralph explain but it can’t be EMI had to be RFI
BE it power or network or anything in between
EMI is magnetic has you thought into how much is needed and how distance makes it become minute in a few feet
tubes can be effected by both but still needs to be close enough if EMI
if there is a router with enabled WiFi even 10 feet away or on another floor or room it can effect things. this is a proven fact and has happened to me
Also most all devices have switching psu now and this includes linaer power supplies
a regulator is also switching as is diodes
So get off the soap box of how you absolutely know it’s your disliked class d amp
Presumably your post was directed toward me? If so, frankly, I'm unsure any of this "stuff" actually addresses any point I raised.
Again, I'm not sure where you're getting this thought from - nor for that matter any of your other thoughts.
I've owned these W4S Class D monoblock amps since 2014 and have thoroughly enjoyed them since their arrival. Why would I keep a pair of monoblocks that I disliked for 11 years running?
Now if only we could listen to a pair of Class D monoblock amps whose designer claims to have absolutely minimized/resolved Class D's bi-directional noise at the source (the cause) - in contrast to consumers like me who at best can only minimize the effects of these distortions.
Last edited:
Not trying to be rude either but I'm still unable to find a single question mark in your collection of thoughts - so I just assumed you were on some soap box yourself.
And my only point of concern in this thread is the bi-directional noise, induced by Class D's high-speed switching modules, being potentially a very real issue (for the OP and his integrated amp) of being shared at the single AC inlet and at the int. amp's linestage that must borrow power from one of the already polluted channels that needs to be addressed.
If you have a specific question or 2 for me, ask away.
No.
Incandescent.
My turn:
1. Ever owned/auditioned a Class D amp in your listening room? Make/model please?
2. Ever owned/auditioned superior bi-directional filtering line conditioners in your listening room? Make/model please?
3. Even utilized both 1 and 2 above at the same time? Please describe.
Judging by your questions, I’m guessing the answers are all no. Had you answered yes to any you most likely would already understand…
1. The valid points I raised about Class D’s bi-directional noise.
2. Your dimmer/lightbulb issue, Class D’s bi-directional noise I mentioned earlier, and numerous other electrical-induced distortions would all be essentially minimized if not rendered non-issues.
That said, if you think you’ve got anything of value to offer the OP’s highly touted Class D integrated and its apparent distortions, I suggest you engage the OP or maybe even Atmasphere directly.
In the meantime, I’m still hoping for someone to post an in-room video of Class D amps whose designer claims, as Atmasphere did, that:
1. Class D’s bi-directional noise is already “pretty low” and well within acceptable industry standards.
2. He’s also solved this apparently already insignificant “pretty low” noise at the source (since, according to him, it’s just RFI).
I can’t imagine any hesitancy producing a video, if there’s any truth to Atmasphere’s claims.
Otherwise, I guess I’ll just have to wait for the breaking headline news stating Class D’s bi-directional noise problem has been eliminated at the source and watch for the soon-to-follow stampede of audiophiles swearing new allegiance to Class D amplification.
Emphasis added: obviously so. But simply because you are unaware of it happening does not mean anything.
Most of this is nonsense.
Its not news that class D amplifier designers have sorted out how to keep switching noise down. We aimed for and achieved 60dB below EC Directives requirements and we were by no means the first to do so. Many tube amplifiers and many class AB solid state amps make more so-called 'bi-directional noise' on the AC line. That noise is generated by rectifiers used in regular power supplies and if its not killed properly in the amp, can radiate at some pretty high frequencies through the air, through grounds and through the AC power.
In our design the rectifier was found to be the main source of noise during testing. We have it snubbed properly (people that know power supply design know what I'm talking about).
No line conditioners needed.
Thanks for your clear answers Atmasphere. I am not sure how I would use the information. I don't see many ways to measure noise without a good desktop scope. And even then, are you measuring the noise that is causing the issue you hear??????
Alrainbow has good advice on isolating devices to see if they are contributors. Its and easy process anyone can apply. You just need to go step by step.
FWIW, to the point of filters, I have measured low order harmonic noise from a variety of filters. 3rd, 5th etc up to 12th and a little more. Many filters add massive pollution to the power lines. I have measured up to 7% thd from very well know filters that is feeding back onto the power line. 2 on one line were creating 15% Thd. I have heard circuit breakers audibly buzz in panels that are connected to well know filters. Unplug the filter and the noise goes away. That does not mean filters don't work. I like filters. I feel most stereo benefit from them.
I have had Atmasphere mono blocks in my system and they were basically dead quiet. They created much less noise than any tube amp I have had in my system. Buzzing out speakers that is. And changing tubes can make a tube amp substantially more noisy or more quiet. I did not put a scope on my power line to see if there were any low level harmonics. I don't have the ability to measure high order noise such as RF or EMF. Not reliably. Only the cheap Tenmar and such type devices that make noise when placed close to a outlet or component. But those are not that useful.
High order, RF type noise is generally not heard as noise. Its a veil that when the noise goes away is lifted. You don't know its there till its gone.
Alrainbow has good advice on isolating devices to see if they are contributors. Its and easy process anyone can apply. You just need to go step by step.
FWIW, to the point of filters, I have measured low order harmonic noise from a variety of filters. 3rd, 5th etc up to 12th and a little more. Many filters add massive pollution to the power lines. I have measured up to 7% thd from very well know filters that is feeding back onto the power line. 2 on one line were creating 15% Thd. I have heard circuit breakers audibly buzz in panels that are connected to well know filters. Unplug the filter and the noise goes away. That does not mean filters don't work. I like filters. I feel most stereo benefit from them.
I have had Atmasphere mono blocks in my system and they were basically dead quiet. They created much less noise than any tube amp I have had in my system. Buzzing out speakers that is. And changing tubes can make a tube amp substantially more noisy or more quiet. I did not put a scope on my power line to see if there were any low level harmonics. I don't have the ability to measure high order noise such as RF or EMF. Not reliably. Only the cheap Tenmar and such type devices that make noise when placed close to a outlet or component. But those are not that useful.
High order, RF type noise is generally not heard as noise. Its a veil that when the noise goes away is lifted. You don't know its there till its gone.
Agreed.
All Class D designers or just you?
Meeting EC directives is good engineering, but as I'm sure you know, compliance rarely translates to silence in real playback systems.
The issue I raised wasn’t about whether a Class D amp measures clean in isolation, but how Class D’s high-frequency switching artifacts behave once two channels share a single AC inlet and in the case of the OP’s int. amp, the linestage must borrow power from an already bi-directional “polluted” channel. That’s the “bi-directional, digital-like noise” many listeners still complain about - perhaps even the OP of this thread. Yet, you haven’t really addressed that – except to say it’s already “pretty low noise” and you corrected it.
Not sure how to respond to this one. You almost make it sound like line conditioners were invented solely for Class D, even though superior (as opposed to inferior) line conditioners can be quite advantageous for 100% analog and Class A and/or B amped systems. And if the line conditioners are bi-directional filtering, they especially benefit when switching artifacts are involved like Class D and digital (DAC, cdp) distortions – both of which happen to be bi-direction.
I’m curious. If you’ve truly solved Class D’s bi-directional noise at the source to the point that any line conditioners aren’t needed, then perhaps your same fix could finally cure the bi-directional digital noise that DACs and CD players are notorious for? Especially since that’s exactly what Class D’s artifacts are often likened to?
Bruno Putzeys says pretty much the same thing I just told you. Any class D designer has to be aware of EU directives and the underlying reasons for their necessity. But I don't doubt some try to fudge things...
It had better! If it does not then its a good bet the CE mark (which can be obtained through self-certification, which means you can lie about it until you get dragged into court...) was not placed on the gear in a legitimate manner.
Right. What you are talking about is hetrodyning, where two noise signals of different frequencies are able to mix and thus generate signals that are the sum of the two and the difference of the two in frequency (the latter of which is likely the main concern). If the two amps are low enough noise then that simply isn't a concern. IMO if placing two modules on the same chassis and feeding them with a common AC cord is a problem, then the modules are simply too noisy regardless of whether they are on a single chassis or not.
I didn't make it seem as anything- the meaning applied is entirely your own.
All I was saying is you don't need a line conditioner with a good class D amp since it is silent, producing less noise than many other kinds of electronics (in particular digital devices). Its obvious you understand that if the switching noise of a class D is able to interfere with digital devices it could cause havoc. So anyone really concerned about such noise will make sure their design has that noise so low its simply not an issue.
Since both are switching technology, yes, paying attention to stray inductances really helps keep DACs and the like low noise as well.
Its worth pointing out here that with any class D amp on the market you can't hear its switching artifacts unless something is dreadfully wrong because the frequency is way too high!!
What you can hear is the distortion that some amps make and sometimes its not pleasant. If the class D amp is not of the self oscillating variety, then you can have a problem with a quantization noise generated by the switching frequency changing due to drift in the local oscillator. This can be heard as a hiss. Unlike the hiss generated by a phono section or the like, this hiss contains lots of artifacts that can be quite fatiguing. So such a class D amp would have to use something like a crystal controlled oscillator where the crystal is in a temperature controlled oven and then its output divided down to the desired switching frequency. But that kind of noise can't be entirely eliminated.
However if you use a self-oscillating design then that problem is neatly side-stepped and the circuitry is actually much simpler.
Such videos are already online.
Please note that the use of 'bidirectional noise' really isn't a thing because if there is noise of any kind at all, its inherently 'bidirectional'; IOW nothing unique to class D or digital operation.
Thanks for the explanation, Ralph. I’m not a designer and I certainly don’t pretend to be one. My comments are derived only from 15 years of living with Class D amps and running a few simple listening experiments over the years – including a few years of selling them and encountering the same bi-directional issues with their installations.
Back around 2007, I recall one doctor client who purchased a pair of nuforce Class D monoblocks and I installed them in his home. He was employing about $7k worth of a popular name brand’s active line conditioners and after about 10 min, I couldn’t take it any more, so I asked if we could take his line conditioners out of the chain. He agreed that it immediately sounded better. So I drove home and grabbed 3 or 4 of my fabulous Foundation Research (now defunct) passive, dedicated, and bi-directional filtering line conditioners and went back to install them. He agreed that his playback presentation was significantly more musical than before and ordered 3 Foundation Research LC-10’s from me.
Come to think of it, I’m unsure I’ve ever heard or read about anybody loving their Class D amps without also employing superior line conditioning. When owners (my customers and on the forums) didn’t employ superior (as opposed to the more popular but inferior type that induce their own sonic harm) line conditioners, it seemed they always complained that Class D was too bright or too detailed and/or there was a digital-like hash, etc, and they always seemed to end up selling them pronto.
Maybe Class D technology has since evolved to the point where such noise is no longer an issue, but thus far I’ve only heard or read such claims from you where you seemed to be in denial stating the noise was already pretty low to begin with and it’s no different than RFI.
And as you just stated earlier that “Its not news that class D amplifier designers have sorted out how to keep switching noise down” to make it sound like all Class D designers had solved this rather significant issue. But again, I’ve only heard such claims from one Class D amp designer – you.
That’s why I keep circling back to system-level audibility rather than compliance numbers or lab behavior. And that’s why my radar remains as high as ever as I can’t help but think you’re in denial that this rather apparent and prominent bi-directional noise was ever an issue.
You’ve given a technical backdrop, but what I’m pointing to is what happens in real playback systems where two channels share the same AC inlet and in this case, the linestage borrows power from an already polluted channel. But I’m also including playback configs where no or inferior line conditioners were employed with any system that included Class D amps. That’s where I and plenty of other listeners have heard the digital-like bi-directional hash or however one wants to label it. So far, I haven’t seen that part addressed directly.
With regard to videos of your Class D amps in action, would you please provide a link to just one that demonstrates your Class D amps’ immunity to this bi-directional noise? I’d love to hear it for myself.
Either way, I’ll leave it here. I’ve shared my own experiences and even provided an in-room video that hopefully shows I’m not just blowing smoke, and the OP can decide for himself what may be negatively impacting his system since installing his Class D integrated — which you claimed must be a malfunction because the problems with Class D bi-directional noise never really existed. Despite the fact many of us have experienced much the same thing in our playback configs for the past 20 years.
I can’t help but think… If the OP’s Class D int. amp is simply a malfunction as you claim here, then it seems to me a lot of perfectly functional Class D amps have been “malfunctioning” much the same way for over two decades.
'Bi-directional' isn't a thing. You are simply talking about switching noise.
Sounds like those amps had a serious problem! I didn't take any class D amps I'd heard seriously back then. It was not until about 8 years ago I heard one that had promise. Turns out that was mostly due to the small sample size to which I was exposed.
Its important when dealing with anecdotes like yours and mine that they suffer from a small sample size, which means that if you draw a conclusion based only on the small sample size, you commit a logical fallacy which will create a blind spot for you (IOW you fool yourself). It is that very reason that prevented us jumping into this technology sooner than we did.
I can point you to a good number of comments and reviews if you like.
That is not what I said. If you look you'll see I said that when you design a class D amp you have to pay attention to stray parasitic inductances in the design, else the unit will radiate noise through the air, through grounds and through the AC power. But I did also say that (paraphrasing) the switching noise is basically RFI as far as other equipment is concerned.
I did point out Bruno Putzeys, several times. On that account the above statement is false.
If you wish to sound credible to anyone who knows about class D or digital design, you will avoid using 'bi-directional'. Its not a thing and it makes it sound as if you have no idea of what you're talking about. I think you do have some idea though, so in the future use 'switching noise'.
Review: Atma-Sphere Class D Amplifiers
Of course you did. You chose the parts and you chose the circuit according to what you count as good. If part specification alone is how we count as 'good' then this is correct. Every time I've seen anyone talk about 'voicing' though, its been in the context of choosing certain parts to...
www.whatsbestforum.com
I think a lot of class D amps have had problems and it goes back far further than just two decades! But they had switching noise problems. Your phrasing here (emphasis added) suggests I said something that I didn't which is pretty much the definition of a strawman logical fallacy.
If you think we didn't listen to our equipment and see how it works in real systems you would be mistaken. We've been making class A triode OTLs for 49 years and we had to be sure that once the class Ds were running right that they also 'sounded' right in comparison to our tube stuff.
The original poster is missing in action so I presume the malfunction was found and corrected.
Republicoftexas69
Well-Known Member
My Ganfet class D amplifiers are dead silent.'Bi-directional' isn't a thing. You are simply talking about switching noise.
Sounds like those amps had a serious problem! I didn't take any class D amps I'd heard seriously back then. It was not until about 8 years ago I heard one that had promise. Turns out that was mostly due to the small sample size to which I was exposed.
Its important when dealing with anecdotes like yours and mine that they suffer from a small sample size, which means that if you draw a conclusion based only on the small sample size, you commit a logical fallacy which will create a blind spot for you (IOW you fool yourself). It is that very reason that prevented us jumping into this technology sooner than we did.
I can point you to a good number of comments and reviews if you like.
That is not what I said. If you look you'll see I said that when you design a class D amp you have to pay attention to stray parasitic inductances in the design, else the unit will radiate noise through the air, through grounds and through the AC power. But I did also say that (paraphrasing) the switching noise is basically RFI as far as other equipment is concerned.
I did point out Bruno Putzeys, several times. On that account the above statement is false.
If you wish to sound credible to anyone who knows about class D or digital design, you will avoid using 'bi-directional'. Its not a thing and it makes it sound as if you have no idea of what you're talking about. I think you do have some idea though, so in the future use 'switching noise'.
Review: Atma-Sphere Class D Amplifiers
Of course you did. You chose the parts and you chose the circuit according to what you count as good. If part specification alone is how we count as 'good' then this is correct. Every time I've seen anyone talk about 'voicing' though, its been in the context of choosing certain parts to...
I think a lot of class D amps have had problems and it goes back far further than just two decades! But they had switching noise problems. Your phrasing here (emphasis added) suggests I said something that I didn't which is pretty much the definition of a strawman logical fallacy.
If you think we didn't listen to our equipment and see how it works in real systems you would be mistaken. We've been making class A triode OTLs for 49 years and we had to be sure that once the class Ds were running right that they also 'sounded' right in comparison to our tube stuff.
The original poster is missing in action so I presume the malfunction was found and corrected.
'Bi-directional' isn't a thing. You are simply talking about switching noise.
Sounds like those amps had a serious problem! I didn't take any class D amps I'd heard seriously back then. It was not until about 8 years ago I heard one that had promise. Turns out that was mostly due to the small sample size to which I was exposed.
Its important when dealing with anecdotes like yours and mine that they suffer from a small sample size, which means that if you draw a conclusion based only on the small sample size, you commit a logical fallacy which will create a blind spot for you (IOW you fool yourself). It is that very reason that prevented us jumping into this technology sooner than we did.
I can point you to a good number of comments and reviews if you like.
That is not what I said. If you look you'll see I said that when you design a class D amp you have to pay attention to stray parasitic inductances in the design, else the unit will radiate noise through the air, through grounds and through the AC power. But I did also say that (paraphrasing) the switching noise is basically RFI as far as other equipment is concerned.
I did point out Bruno Putzeys, several times. On that account the above statement is false.
If you wish to sound credible to anyone who knows about class D or digital design, you will avoid using 'bi-directional'. Its not a thing and it makes it sound as if you have no idea of what you're talking about. I think you do have some idea though, so in the future use 'switching noise'.
Review: Atma-Sphere Class D Amplifiers
Of course you did. You chose the parts and you chose the circuit according to what you count as good. If part specification alone is how we count as 'good' then this is correct. Every time I've seen anyone talk about 'voicing' though, its been in the context of choosing certain parts to...
I think a lot of class D amps have had problems and it goes back far further than just two decades! But they had switching noise problems. Your phrasing here (emphasis added) suggests I said something that I didn't which is pretty much the definition of a strawman logical fallacy.
If you think we didn't listen to our equipment and see how it works in real systems you would be mistaken. We've been making class A triode OTLs for 49 years and we had to be sure that once the class Ds were running right that they also 'sounded' right in comparison to our tube stuff.
The original poster is missing in action so I presume the malfunction was found and corrected.
Thanks for the link, Ralph. I hadn’t planned on returning to this thread, but since you actually posted a video of your amps in action, I’m compelled to respond.
Coincidentally, my July 6 in-room video happens to be of a different pressing and artist, but close enough for some side-by-side comparisons.
Granted, we all discern/interpret what we hear differently. But personally, I’m hearing some overly bright, overly detailed, thin/lean, slightly compressed, congested, boxy/roomy bass, and a touch of digital-like hash in one of these videos — very similar to traits Class D and even digital owners have been complaining about for 20+ years. But I’ll leave it to you and others to decide which.
Normally I’d suggest two notches below max volume for MacBook + headphone listeners, but in this case, to cut to the chase, I’d suggest max volume for both.
Either way, I think this turned into a pretty good accidental illustration of what superior line conditioning can bring to the table – especially those line conditioners with bi-directional filtering capabilities to address bi-directional digital-like distortions like those annoying "malfunctions" induced by Class D amps and anything digital. Shoot, this may even be a pretty good accidental illustration between analog and digital too?
And to be clear, I’m not out to convince anyone who’s already certain they don’t hear these things — this is just for the OP, if he's still around, and anybody else curious enough to judge with their own ears.
Last edited:
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |